Abstracts

Rationale: Photoplethysmography (PPG) is a non-invasive optical technique for measuring variations in blood perfusion in peripheral tissues. PPG changes may assist in the detection of generalized tonic clonic seizures and may correlate with biomarkers of SUDEP, such as post-ictal EEG suppression. Methods: We asked patients to wear a portable wristband sensor that records PPG signals on wrists or ankles during audio/video EEG long term monitoring. Findings were correlated with seizures detected by video EEG monitoring. We analyzed 3 time periods of the PPG signal: baseline period before the seizure onset; pre-seizure phase as the time period immediately preceding seizure onset, and the post-seizure phase after seizure termination. Signals were pre-processed (band-pass filtering [0.1-8] Hz), analyzed in MATLAB software, and signal feature analysis was performed. All the analyses for individual PPG signal features were carried out using random-intercept to account for variability between subjects and with-in subject correlations. Results: We prospectively enrolled 126 patients admitted to the epilepsy monitoring unit at Boston Children's Hospital between February 2015 and March 2016. Out of 126 patients recorded, we analyzed 7 patients who had a generalized tonic clonic (GTC) seizure during the recording, and we analyzed PPG graphs of these patients. The age of patients ranged from 11 to 17 years old with the mean of14.1 years. Age of epilepsy onset, epilepsy type, EEG findings and MRI features are depicted in the table 1. Comparison of the baseline with both pre-seizure and post-seizure signals showed significant results in pairwise comparisons of the following features: Frequency, Peak amplitude, Duration, Increasing slope, Decreasing slope, Smoothness and Area under the curve (table 2). All of the aforementioned features, changed significantly from baseline to the immediate pre-seizure interval, and also from baseline to post-seizure periods. Namely, our data shows that there is significant change from baseline in frequency, peak amplitude, slope, smoothness and area under the curve for PPG signals, before and after the seizure occurrence. Conclusions: In patients with epilepsy, monitoring of PPG signals may be helpful in detecting seizures. Findings of this study encourage further data collection and validation through implementation of this method in future wearable devices for seizure detection and prediction in epilepsy patients. Funding: Epilepsy Research Fund